Hmm, on second thought, I can't quite seem to get this to work. Trying the example from __test_matthias_question line for line, it throws a TypeError on classifier initialization.<div><br><div><div>TypeError: Unexpected keyword argument transformer=<ufunc 'absolute'> for BoostedClassifierSensitivityAnalyzer: 4 ca:ca{base_sensitivities raw_results null_t null_prob+}. Valid parameters are ['base_sensitivities', 'raw_results', 'null_t', 'null_prob']</div>
<br></div><div>wtf, right? </div><div><div>Current date: 2010-11-01 17:43</div><div>PyMVPA:</div><div> Version: 0.5.0.dev</div><div> Path: /software/python/nipype0.3/lib/python2.6/site-packages/pymvpa-0.5.0.dev-py2.6-linux-x86_64.egg/mvpa/__init__.pyc</div>
<div> Version control (GIT):</div><div> GIT information could not be obtained due "/software/python/nipype0.3/lib/python2.6/site-packages/pymvpa-0.5.0.dev-py2.6-linux-x86_64.egg/mvpa/.. is not under GIT"</div><div>
SYSTEM:</div><div> OS: posix Linux 2.6.31-22-generic #65-Ubuntu SMP Thu Sep 16 16:21:34 UTC 2010</div><div> Distribution: Ubuntu/9.10/karmic</div><div>EXTERNALS:</div><div> Present: cPickle, ctypes, good scipy.stats.rdist, good scipy.stats.rv_discrete.ppf, griddata, gzip, libsvm, libsvm verbosity control, lxml, matplotlib, nibabel, nose, numpy, pylab, pylab plottable, pywt, pywt wp reconstruct, running ipython env, scipy</div>
<div> Absent: atlas_fsl, atlas_pymvpa, elasticnet, glmnet, h5py, hcluster, lars, mdp, mdp ge 2.4, nifti, nifti ge 0.20090205.1, nipy, openopt, pprocess, pywt wp reconstruct fixed, reportlab, rpy2, sg ge 0.6.4, sg ge 0.6.5, sg_fixedcachesize, shogun, shogun.krr, shogun.lightsvm, shogun.mpd, shogun.svrlight, weave</div>
<div> Versions of critical externals:</div><div> ipython : 0.10.1</div><div> scipy : 0.7.2</div><div> numpy : 1.4.1</div><div> ctypes : 1.1.0</div><div> matplotlib : 1.0.0</div><div> lxml : 2.3.beta1</div>
<div> numpy : 1.4.1</div><div> pywt : 0.2.0</div><div> Matplotlib backend: TkAgg</div><div>RUNTIME:</div><div> PyMVPA Environment Variables:</div><div> PyMVPA Runtime Configuration:</div><div> [externals]</div>
<div> have griddata = yes</div><div> have pprocess = no</div><div> have good scipy.stats.rdist = yes</div><div> have pylab plottable = yes</div><div> have pywt wp reconstruct = yes</div><div> have mdp = no</div><div>
have lxml = yes</div><div> have running ipython env = yes</div><div> have nibabel = yes</div><div> have sg_fixedcachesize = no</div><div> have elasticnet = no</div><div> have shogun.mpd = no</div><div> have matplotlib = yes</div>
<div> have pywt wp reconstruct fixed = no</div><div> have scipy = yes</div><div> have reportlab = no</div><div> have openopt = no</div><div> have libsvm = yes</div><div> have h5py = no</div><div> have shogun.krr = no</div>
<div> have nifti ge 0.20090205.1 = no</div><div> have nose = yes</div><div> have weave = no</div><div> have atlas_fsl = no</div><div> have ctypes = yes</div><div> have hcluster = no</div><div> have sg ge 0.6.4 = no</div>
<div> have sg ge 0.6.5 = no</div><div> have good scipy.stats.rv_discrete.ppf = yes</div><div> have libsvm verbosity control = yes</div><div> have mdp ge 2.4 = no</div><div> have shogun.svrlight = no</div><div> have shogun = no</div>
<div> have nipy = no</div><div> have glmnet = no</div><div> have lars = no</div><div> have nifti = no</div><div> have atlas_pymvpa = no</div><div> have cpickle = yes</div><div> have numpy = yes</div><div> have pylab = yes</div>
<div> have rpy2 = no</div><div> have shogun.lightsvm = no</div><div> have pywt = yes</div><div> have gzip = yes</div><div> </div><div> [general]</div><div> verbose = 1</div><div> Process Information:</div><div> Name: ipython</div>
<div> State: R (running)</div><div> Tgid: 18461</div><div> Pid: 18461</div><div> PPid: 12889</div><div> TracerPid: 0</div><div> Uid: 85302 85302 85302 85302</div><div> Gid: 33368 33368 33368 33368</div>
<div> FDSize: 256</div><div> Groups: 1007 1008 33368 33376 68218 68236 68252 777333 </div><div> VmPeak: 372540 kB</div><div> VmSize: 372512 kB</div><div> VmLck: 0 kB</div><div>
VmHWM: 64688 kB</div><div> VmRSS: 64684 kB</div><div> VmData: 58408 kB</div><div> VmStk: 244 kB</div><div> VmExe: 2108 kB</div><div> VmLib: 53604 kB</div>
<div> VmPTE: 700 kB</div><div> Threads: 2</div><div> SigQ: 0/16382</div><div> SigPnd: 0000000000000000</div><div> ShdPnd: 0000000000000000</div><div> SigBlk: 0000000000000000</div>
<div> SigIgn: 0000000001001000</div><div> SigCgt: 0000000180000002</div><div> CapInh: 0000000000000000</div><div> CapPrm: 0000000000000000</div><div> CapEff: 0000000000000000</div><div>
CapBnd: ffffffffffffffff</div>
<div> Cpus_allowed: ff</div><div> Cpus_allowed_list: 0-7</div><div> Mems_allowed: 00000000,00000001</div><div> Mems_allowed_list: 0</div><div> voluntary_ctxt_switches: 2295</div><div> nonvoluntary_ctxt_switches: 33</div>
</div><div><br></div><div><br></div><div><div class="gmail_quote">On Mon, Nov 1, 2010 at 2:10 PM, Michael Waskom <span dir="ltr"><<a href="mailto:mwaskom@mit.edu">mwaskom@mit.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Hi Yarik,<div><br></div><div>Thanks! The pointer to that test is very helpful indeed.</div><div><br></div><div>Best,</div><div>Mike<div><div></div><div class="h5"><br><br><div class="gmail_quote">On Mon, Nov 1, 2010 at 2:02 PM, Yaroslav Halchenko <span dir="ltr"><<a href="mailto:yoh@dartmouth.edu" target="_blank">yoh@dartmouth.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Mike,<br>
<br>
sorry that it causes you troble... but RFE is indeed somewhat difficult<br>
to figure out...<br>
<br>
but before anything:<br>
have you seen<br>
<a href="http://dev.pymvpa.org/featsel.html#recursive-feature-elimination" target="_blank">http://dev.pymvpa.org/featsel.html#recursive-feature-elimination</a><br>
<br>
also look may be at<br>
mvpa/tests/test_rfe.py: def __test_matthias_question(self):<br>
which has one of RFE structures<br>
<br>
in general classifier with RFE as feature selection should be treated as<br>
any other classifier<br>
<br>
also -- do you really think that RFE is necessary in your case? may be<br>
some simpler feature selection or SMLR would do?<br>
<br>
P.S. it would be of greater benefit to everyone if we continue this<br>
communication on the mailing list, if you don't mind of cause<br>
<div><div></div><div><br>
On Mon, 01 Nov 2010, Michael Waskom wrote:<br>
<br>
> Hey Yarik,<br>
<br>
> I'm trying to figure out how to use the RFE class in pymvpa 0.5dev.<br>
> Sadly, there doesn't seem to be much in the (otherwise<br>
> super-helpful) documentation about it. I get all of the parameters,<br>
> but I'm not quite sure how it might fit into a typical analysis, and<br>
> I'm very interested in trying out some recursive feature selection. Do<br>
> you have a simple use case you could put up as a gist (or elsewhere) so<br>
> I can get the basic idea? Nothing too didactic is required.<br>
<br>
> Thanks!<br>
<br>
> Mike<br>
<br>
</div></div><font color="#888888">--<br>
Yaroslav O. Halchenko<br>
Postdoctoral Fellow, Department of Psychological and Brain Sciences<br>
Dartmouth College, 419 Moore Hall, Hinman Box 6207, Hanover, NH 03755<br>
Phone: +1 (603) 646-9834 Fax: +1 (603) 646-1419<br>
WWW: <a href="http://www.linkedin.com/in/yarik" target="_blank">http://www.linkedin.com/in/yarik</a><br>
</font></blockquote></div><br></div></div></div>
</blockquote></div><br></div></div>